Ok, so it was just two days ago when I tweeted this:
here's your DM rates mini-tantrum in action in stock-bond return correlations… pic.twitter.com/UPDf7DWPLD
— Walter White (@heisenbergrpt) July 18, 2017
The point was that the mini-tantrum in DM rates that Mario Draghi kicked off in Sintra, Portugal on June 27 was accompanied by sharply less negative stock/bond return correlations.
In other words, the extent to which equities were doing their part to hedge the bond selloff was diminishing. Quite a few folks suspect that’s what helped throw CTAs and risk parity for a loop, with the former turning in their worst two-week performance as a group in a decade earlier this month.
“During the week of June 26, when the bond sell-off started to weigh on risky assets, the broad SG/Newedge index, with the top 20 CTAs globally, saw the largest weekly drawdown since the GFC (Exhibit 15),” Goldman writes, in a note out Wednesday afternoon.
“Risk parity strategies have also suffered – albeit less compared with their history (Exhibit 16) – although they remain at risk from rates volatility spilling over to risky assets, as they have larger allocation to bonds compared with traditional balanced portfolios and the current low regime has likely driven higher allocations to risky assets.”
The message there was simple: a rise in rates vol. puts everything in jeopardy and the first place you’re going to see the pain is in systematic strats, which is ironic because that very same pain will likely cause those very same strats to deleverage into a falling market, thus catalyzing an even deeper selloff. (And God help us all if inverse and levered VIX ETPs are forced to panic buy VIX futs in the middle of the whole thing)
This is the subject of the above-mentioned new note from Goldman, appropriately entitled “Goldilocks and the Three Central Banks.”
Note the chart in the tweet embedded above. So that’s stock/bond return correlations. Obviously, you want that to be negative if one or the other is selling off. That’s the whole idea behind holding both stocks and bonds for diversification. Put differently, you want the equity/bond yield correlation to be positive (bond yields rise i.e. bonds selloff, stocks rise).
Part and parcel of this whole conversation is whether stocks perceive a rise in yields as a good sign (i.e. as an encouraging sign that the ever elusive “reflation” narrative is actually playing out), or as a bad sign (i.e. if stocks sense a policy mistake and we get a “tantrum” episode). In the post-crisis era, the rates threshold beyond which stocks stop looking at higher yields through rose colored glasses seems to have fallen materially, meaning the tantrum point is now lower.
So far, equities have held up, but as the chart shown here at the outside suggests, things are getting dicey and the same is true in credit, where the most bulletproof of all assets is being tested.
Here’s Goldman (again, in Exhibit 3 below, you’re looking at equity/bond yield correlations, so it’s flipped versus the chart in the tweet):
Following more hawkish surprises from central banks in June, global bond yields have moved sharply higher, signaling the potential end of the ‘Goldilocks scenario’.
Risky assets digested the increase in bond yields only reasonably well – 3-month equity/bond yield correlations stayed positive (Exhibit 3) and credit spreads buffered part of the increases (Exhibit 4). But correlations are starting to reverse: since June 26, when the bond sell-off started, they have reversed further. Particularly in Europe, both equities and HY credit have suffered from higher yields, likely reflecting concerns over tighter ECB policy and a stronger Euro.
Equity/bond yield correlations have shifted positive since the late 1990s, with anchored inflation and more market-friendly central banks – this remains our base case going forward. As a result, from low levels higher bond yields are often seen as a positive signal that deflation risk is fading and growth is picking up (rather than inflation overshooting). Assuming the growth/rates mix is favourable and yields increase gradually, equities can buffer bond yields from elevated levels of equity risk premia.
Clearly, the worry is that we could get a rates shock far more severe than what we saw in the wake of Sintra.
That harkens back to what we said last week. Namely this:
To be sure, there’s something comical about the juxtaposition between this idea that DM central bankers are trying their best to “carefully” telegraph an exit from crisis-era policies and the hair-on-fire, mad dash to coordinate a hawkish message that we’ve witnessed since Draghi’s comments in Portugal on June 27.
As noted, they’re trying to micromanage this by the hour (that’s the “careful” telegraphing) but really, you need only look at a one-month chart of DM bond yields to see why it’s tempting to call this a rather ham-handed effort:
So far, that hasn’t translated into the kind of equity weakness you might imagine would accompany a “tantrum” but the jury is still out – remember, this has only been going on for three weeks. That is, try to extrapolate from that chart what yields are going to look like in a year if they don’t figure out how to get the messaging “right.”
“The risk of ‘rate shocks’ remains from current low levels of bond yields, e.g. due to a strong unexpected pick-up in inflation or a monetary policy error,” Goldman goes on to warn, adding that “in periods of negative rate shocks, like the US ‘taper tantrum’ in May 2013 (Exhibit 11), risky assets unsurprisingly suffer.”
And guess what the gasoline on the fire will be?
That’s right: the very same programmatic strats that will themselves suffer the most in a tantrum scenario. Here’s Goldman one more time:
As we discussed before, in the event of a sharp and sustained increase in bond yields and volatility, momentum investors such as CTAs, risk parity and vol target investors are likely to suffer and could exacerbate volatility in the near term due to their positioning.